This crate handles a group of threads (whose closures have the same return type) as one unit,
letting you join()
or join_timeout()
on the first thread of the group that is ready. Internally,
it uses a channel to notify about finished threads.
This is useful when you want to check whether any one of your threads has panicked or returned
early, for error-handling or for progress report. This isn't possible with
std::thread::JoinHanlde.join()
because it is a blocking call, so thread 2 might panic or finish
while you're waiting indefinitely on thread 1.
If ThreadGroup
isn't what you needed after all and can't be improved to suit your needs, have a
look at these other crates :
- Thread_tryjoin gives the same benefit as
ThreadGroup.join_timeout()
with an API that is closer tostd
's. But it depends on a Linux-only API, and only handles one thread at a time (so thattry_join()
ing a lot of thread will end up wasting either time or CPU). - Thread-control and Runloop let you stop threads, and test without blocking if they panicked or finished. Again, this only handles one thread at a time.
- Rayon is a higher-level abstraction to spread a computation over multiple threads.
- Thread pools are a common idiom to handle short computations without repeatedly paying the high setup cost.
- Future isn't related to threads but is often a better way to handle "start task B whenever task A finishes" algorythms.
Please create issues and send pull request via Github.
Threadgroup is licensed as MIT. Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the MIT license, shall be licensed as above, without any additional terms or conditions.